Electrospray mass spectrometry is a powerful analytical tool that has been broadly applied to bio-molecular structure analysis (i.e., Proteins, Peptides and DNA). See Electrospray Ionization Mass Spectrometry Fundamentals, Instruments, and Applications, Richard B. Cole, John Wiley and Sons, 1997. This technique plays a central role in the development of most pharmaceutical drugs and is being used to perform quantitative measurement of human exposure to carcinogens. Because of the size and potential revenues of the pharmaceutical market, there is interest in developing instrumentation based on, and technical enhancements to, electrospray mass spectrometry.
In recent years there has been a general trend to minimize the amount of sample required for analysis and micro-electrospray ionization (micro-ESI, micro-ES) and nanospray describe two of these approaches. These two methods share a lot in common, and they are often used interchangeably. Micro-ES is a miniaturized electrospray source with the same system components as “conventional” electrospray. These include a source of pumped liquid flow containing the sample for analysis, a small diameter sharp hollow needle through which liquid is pumped, and a source of high voltage to generate the spray. Nanospray relies on the electrostatic attraction of the liquid inside the needle towards an attractor counter-electrode to generate the flow rather than a pump. This characteristic makes nanospray very attractive as a means to minimize sample waste. Since electrospray, micro-ES, and nanospray are all species of a generic class referred to as electrospray they will be interchangeably referred to as electrospray in this patent.
The nature of the electrospray ionization process makes sample preparation a major consideration. The presence of solvent and buffer salts along with the sample significantly increases spectral complexity and degrades detection limits. The electrospray ionization process produces an abundance of solvent ions that give an intense mass spectral background that can severely limit identification of many compounds at trace levels in solution. Even without the solvent ions to contend with, many applications require working with complex mixtures that necessitate some degree of separation prior to mass analysis. See J. Lee, J. F. Kelly, I. Chernushevich, D. J. Harrison, and P. Thibalut “Separation and Identification of Peptides from Gel-Isolated Membrane Proteins Using a Microfabricated Device for Combined Capillary Electrophoresis/Nanoelectrospray Mass Spectrometry,” Anal. Chem. 2000, 72,599–609. Better methods for elimination of unwanted solvent and separation of sample ions from background are therefore needed.
Electrospray mass spectrometry (ES-MS) provides a powerful tool for structure determination of peptides, proteins. This is important, as structure to a large extent defines the function of the protein. The structural information about a protein is typically determined from its amino acid sequence. To identify the sequence, the protein is usually digested by enzymes, and the peptide fragments are sequenced by tandem mass spectrometry. Another possible way to obtain the sequence is to digest the protein and measure the molecular weights of the peptide fragments. These are the input data for a computer program which digests theoretically all the proteins being found in the data base and the theoretical fragments are compared with the measured molecular weights.
Recently, it has been noticed that Ion Mobility Spectrometry can provide useful information to an electrospray/mass-spectrometry measurement. Ion Mobility spectrometry is ordinarily an atmospheric pressure technique which is highly sensitive to the shape and size of a molecule. Protein identification thorough the combination of an IMS and mass spectrometer may eliminate the need for protein digestion, simplifying sample preparation.
Commercially available IMS systems are based on time-of-flight (TOF), i.e., they measure the time it takes ions to travel from a shutter-gate to a detector through an inert atmosphere (1 to 760 Torr.). The drift time is dependent on the mobility of the ion (i.e., its size, mass and charge) and is characteristic of the ion species detected. TOF-IMS is a technique useful for the detection of many compounds including narcotics, explosives, and chemical warfare agents. See PCT Application Ser. No. PCT/CA99/00715 incorporated herein by this reference and U.S. Pat. No. 5,420,424 also incorporated herein by this reference. In ion mobility spectrometry, gas-phase ion mobility is determined using a drift tube with a constant low field strength electric field. Ions are gated into the drift tube and are subsequently separated based on differences in their drift velocity. The ion drift velocity under these conditions is proportional to the electric field strength and the ion mobility, which is determined from experimentation, is independent of the applied field. Current spectrometers use conventionally machined drift tubes (minimum size about 40 cm3) for ion identification.
In conventional time-of-flight ion mobility spectrometers (TOF-IMS) ion identification is done in a low strength electric field (less than 1000 V/cm) where the coefficient of mobility for each ion is essentially independent of field strength [.W. McDaniel and Edward A. Mason, The mobility and diffusion of ions in gases, John Wiley & Sons, 1973].
At high electric fields, ion mobility becomes dependent upon the applied electric field strength and the ion drift velocity may no longer behave linearly with field strength. This principle is utilized in the subject of this disclosure.
The field asymmetric waveform ion mobility spectrometer (FAIMS, also known as RF-IMS) utilizes these significantly higher electric fields, and identifies the ion species based on the difference in its mobility in high and low strength electric fields.
The FAIMS spectrometer uses an ionization source, such as an ultra violet photo-ionization lamp, to convert a gas sample into a mixture of ion species with each ion type corresponding to a particular chemical in the gas sample. The ion species are then passed through an ion filter where particular electric fields are applied between electrodes to select an ion type allowed to pass through the filter. Once through the filter the ion type hits a detector electrode and produces an electrical signal. To detect a mixture of ion species in the sample, the electric fields applied between the filter electrodes can be scanned over a range and a spectrum generated. The ion filtering is achieved through the combination of two electric fields generated between the ion filter electrodes, an asymmetric, periodic, radio frequency (RF) electric field, and a dc compensation electric field. The asymmetric RF field has a significant difference between its peak positive field strength and negative field strength. The asymmetric RF field scatters the ions and causes them to deflect to the ion filter electrodes where they are neutralized, while the compensation field prevents the scattering of a particular ion allowing it to pass through to the detector. The ions are filtered in instruments on the basis of the difference in the mobility of the ion at high electric fields relative to its mobility at low electric fields. That is, the ions are separated due to the compound dependent behavior of their mobility at high electric fields relative to their mobility at low electric fields.
The FAIMS approach is based on an observation of Mason and McDaniel [.W. McDaniel and Edward A. Mason, The mobility and diffusion of ions in gases, John Wiley & Sons, 1973] who found that the mobility of an ion is affected by the applied electric field strength. Above an electric field to gas density ratio (E/N) of 40 Td (E>10,700V/cm at atmospheric pressure) the mobility coefficient K(E) has a non-linear dependence on the field. This dependence is believed to be specific for each ion species. Below are some examples from Mason and McDaniel [.W. McDaniel and Edward A. Mason, The mobility and diffusion of ions in gases, John Wiley & Sons, 1973]. The mobility for the cluster ion CO+CO increases with increasing field strength (FIG. 7-1-K-1 in reference [.W. McDaniel and Edward A. Mason, The mobility and diffusion of ions in gases, John Wiley & Sons, 1973]). For some molecular and atomic ions the coefficient of mobility can change in a more complex way. For example, for atomic ions K+, the mobility coefficient in carbon monoxide gas increases with increasing field by as much as 20%, but above E/N˜200 Td the coefficient starts to decrease (FIG. 7-1-K-3 in reference [.W. McDaniel and Edward A. Mason, The mobility and diffusion of ions in gases, John Wiley & Sons, 1973]). For some other ions for example N+, N3+ and N4+ the mobility changes very little (FIG. 7-1-H-1/2 in reference [.W. McDaniel and Edward A. Mason, The mobility and diffusion of ions in gases, John Wiley & Sons, 1973]). FIG. 1A illustrates schematically three possible ion mobility dependencies on electric field. For simplicity we will assume that the low field value of the mobility K(Emin) in a weak electric field (E approximately 102–103 V/cm) is the same for all three ion types. However, at Emax the value of the mobility coefficient K(Emax) is different for each ion type.
The field dependence of the mobility coefficient K(E) can be represented by a series expansion of even powers of E/N [18]K(E)=K(0)[1+α1(E/N)2+α2(E/N)4+ . . . ]  (1)where K(0) is the coefficient of mobility of the ion in a weak electric field, and α1, α2 are coefficients of the expansion. This equation can be simplified by using an effective α(E) as shown in equation 2 [T. W. Carr, Plasma Chromatography, Plenum Press, New York and London, 1984],K(E)≈K(0)[1+α(E)].  (2)According to this expression when α(E)>0 the mobility coefficient K(E) increases with field strength, when α(E)˜0 the mobility K(E) does not change, and when α(E)<0 then K(E) decreases with increasing field strength. An expression for the field dependent mobility coefficient can also be derived from momentum and energy balance considerations. Where the energy of the ion ε=3/2 kTeff can be expressed as a function of its effective temperature [18–20].
                              K          ⁡                      (            E            )                          =                              v            E                    =                                    q              N                        ⁢                                          (                                  1                                      3                    ⁢                    μ                    ⁢                                                                                  ⁢                                          kT                      eff                                                                      )                                            1                /                2                                      ⁢                                          1                                  Ω                  ⁡                                      (                                          T                      eff                                        )                                                              .                                                          (        3        )            
The case where α(E)<0 can be explained based on the model presented in equation 3, if one assumes the value of the ion neutral cross-section Ω(Teff) does not change significantly for rigid-sphere interactions [T. W. Carr, Plasma Chromatography, Plenum Press, New York and London, 1984, E. A. Mason and E. W. McDaniel, Transport Properties of Ions in Gases, Wiley, New York, 1988] and the reduced mass μ is constant. Under these conditions one finds that the mobility K(E) will decrease if the effective temperature, or energy, of the ion increases. Physically this effect has a simple explanation. When the electric field strength is increased the ions are driven harder through the neutral gas. This increases the ion neutral collision frequency, which leads to a reduced average ion velocity and a reduced ion mobility coefficient.
The rigid-sphere model however, does not explain the experimental results which show that with certain ions the mobility increases with increasing electric field (α(E)>0). One of the possible explanations for the increased mobility at elevated values of E/N is offered when one allows for ion de-clustering at high field strengths to occur. Ions in ambient conditions in a weak electric field generally do not exist in a free state. They are usually in cluster form (for example, MH+(H2O)n) with n polar molecules such as water attached. As the electric field strength is increased the kinetic energy and consequently the effective temperature (Teff) of the ion increases due to the energy imparted between collisions. This can lead to a reduction in the level of ion clustering (reduction in n) resulting in a smaller ion cross-section Ω(Teff) and a smaller reduced mass μ for the ion. According to equation 3 then, if do to de-clustering the cross-section and reduced mass decrease in a sufficient manner to offset the increase in Teff the case where α(E)>0 can be explained.
The third case when α(E)˜0 can be explained by a decrease in ion cross section due to de-clustering which is offset by an increase in the effective temperature of the ion. This results in no net change to the mobility coefficient of the ion.
The mechanism of operation of the FAIMS for ion filtering is described in the following. Consider three kinds of ions with different mobility coefficient dependencies on electric field (i.e., α(E)>0, α(E)<0, α(E)˜0) which are formed, due to local ionization of neutral molecules, at the same location in a narrow gap between two electrodes, as shown on FIG. 1B. A stream of carrier gas transports these ions longitudinally down the drift tube between the gap. If an asymmetric RF electric field is then applied to the electrodes the ions will oscillate in a perpendicular direction to the carrier gas flow, in response to the RF electric field, while moving down the drift tube with the carrier gas. A simplified asymmetric RF electric field waveform (FIG. 1C) with maximum field strength |Emax|>10,000 V/cm and minimum field strength |Emin|<<|Emax| is used here to illustrate the operation principle of the RF-IMS. The asymmetric RF waveform is designed such that the time average electric field is zero and|Emax|t1=|Emin|t2=β.  (1)t1 is the portion of the period where the high field is applied and t2 is the time the low field is applied. β is a constant corresponding to the area under-the-curve in the high field and low field portions of the period. The ion velocities in the y-direction are given byVy=K(E)E(t).  (2)Here K is the coefficient of ion mobility for the ion species and E is the electric field intensity, in this case entirely in the y-direction. If the amplitude of the positive polarity RF voltage pulse (during t1) produces an electric field of strength greater than 10,000V/cm then the velocity towards the top electrodeVup=Kup|Emax|  (3)will differ for each of the ion species (FIG. 1B) since, as shown in FIG. 1A, the coefficient of mobility Kup for each ion at the high field condition is different. The ions with α(E)>0 will move faster and ions with α(E)<0 will have the smallest velocity, therefore, the slope of each ion's trajectory will also differ. In the next portion of the period (t2), once the polarity of the RF field has switched, all three ion types will begin moving with the same velocityVdown=K(Emin)|Emin|  (4)down towards the bottom plate. In this low field strength condition (see FIG. 1A) all three ion types will have the same mobility coefficient Kdown. Therefore, all three ion trajectories will have the same slope in this portion of the period (FIG. 1B).
The ion displacement from its initial position in the y-direction is the ion velocity in the y-direction Vy multiplied by the length of time Δt the field is appliedΔy=VyΔt.  (5)In one period of the applied RF field the ion moves in both the positive and negative y-directions. By substituting equation 2 into equation 5 the average displacement of the ion over one period of the RF field can be written asΔyRF=Kup|Emax|t1−Kdown|Emin|t2.  (6)Using equation 1 this expression can be re-written asΔyRF=β(Kup−Kdown)=βΔK.  (7)Since β is a constant determined by the applied RF field, the y-displacement of the ion per period of the RF field T=t1+t2 depends on the change in mobility of the ion between its high and low field conditions. Assuming the carrier gas only transports the ion in the z-direction. The total ion displacement Y (in the y-direction) from its initial position (due to the electric field) during the ions residence time tres between the ion filter plates can be expressed as
                    Y        =                                                            Δ                ⁢                                                                  ⁢                                  y                  RF                                                            (                                                      t                    1                                    +                                      t                    2                                                  )                                      ⁢                          t              res                                =                                                    β                ⁢                                                                  ⁢                Δ                ⁢                                                                  ⁢                K                            T                        ⁢                          t              res                                                          (        8        )            The average ion residence time inside the ion filter region is given in equation 9. A is the cross-section area of the filter region, L is the length of the ion filter electrodes, V is the volume of the ion filter region V=AL, and Q is the volume flow rate of the carrier gas.
                              t          res                =                              AL            Q                    =                                    V              Q                        .                                              (        9        )            Substituting equation 9 into equation 8, noting from equation 1 that β=|Emax|t1 and defining the duty cycle of the RF pulses as D=t1/T. The equation for displacement of the ion species, equation 8, can be re-written as
                    Y        =                              Δ            ⁢                                                  ⁢                          KE              max                        ⁢            VD                    Q                                    (        10        )            where Y is now the total displacement of the ion in the y-direction based on the average ion residence time in the ion filter region. From equation 10 it is evident that the vertical displacement of the ions in the gap are proportional to the difference in coefficient of mobility between the low and high field strength conditions. Different species of ions with different ΔK values will displace to different values of Y for a given tres. All the other parameters including the value of the maximum electric field, the volume of the ion filter region, the duty cycle and the flow rate, to first order are essentially the same for all ion species.
When a low strength DC field (|Ec|<|Emin|<<|Emax|) is applied in addition to the RF field, in a direction opposite to the average RF-induced (y-directed) motion of the ion, the trajectory of a particular ion species can be “straightened”, see FIGS. 1D(1), 1D(2), 1D(3). This allows the ions of a particular species to pass unhindered between the ion filter electrodes while ions of all other species are deflected into the filter electrodes. The DC voltage that “tunes” the filter and produces a field which compensates for the RF-induced motion is characteristic of the ion species and is called the compensation voltage. A complete spectrum for the ions in the gas sample can be obtained by ramping or sweeping the DC compensation voltage applied to the filter. The ion current versus the value of the sweeping voltage forms the RF-IMS spectra. If instead of sweeping the voltage applied to one of the ion filter electrodes, a fixed DC voltage (compensation voltage) is applied, the spectrometer will work as continuous ion filter allowing only one type of ion through.
In PCT Application Ser. No. PCT/CA99/00715, an electrospray ionization chamber or electrospray source is used to create ions which are ultimately transported to an analytical region which is subject to both a high frequency voltage asymmetric waveform and a DC offset voltage.
It is therefore an object of the present invention to provide method and apparatus for improved detection of compounds using field asymmetric waveform ion mobility.